Pacific water shrew (Sorex bendirii) COSEWIC assessment and status report: chapter 7

Search effort

Most of the occurrence records for this species in Canada are based on incidental captures or observations by museum collectors and naturalists or recent surveys with pitfall traps (Craig and Vennesland 2004b) conducted in very limited geographic areas as part of environmental assessments. However, two intensive surveys have been done for this species in Canada that can be used to assess presence-absence and sampling effort. From 1989-1991, Seip (unpublished data) and Seip and Savard (1990) sampled 22 sites in clear-cut, second growth, and old growth forest stands within the range of the Pacific water shrew in the Capilano, Seymour, and Coquitlam rivers’ watersheds in the southern Coast Mountains (Figure 5). They employed pitfall traps consisting of cans, Museum Special snap traps and Sherman live traps with a sampling effort of 17,315 pitfall trap nights, 6,054 snap trap nights and 4,791 Sherman live trap nights. Only 5 Pacific water shrews were taken at 2 of 22 sites during this sampling period. All were taken in pitfall traps. Seip and Savard (1990) give no information on the proximity of their traps to water or riparian habitats.

Figure 5. Presence-absence data for Pacific water shrews (Sorex bendirii) at 61 locations surveyed within its range. Based on 22 locations surveyed from 1989-1992 by Seip (unpublished data) and Seip and Savard (1990) and 39 locations sampled in 1992 by Zuleta and Galindo-Leal (1994).

Figure 5. Presence-absence data for Pacific water shrews (Sorex bendirii) at 61 locations surveyed within its range. Based on 22 locations surveyed from 1989-1992 by Seip (unpublished data) and Seip and Savard (1990) and 39 locations sampled in 1992 by Zuleta and Galindo-Leal (1994).

In 1992, Zuleta and Galindo-Leal (1994) surveyed 55 sites at 39 locations (Figure 5) for Pacific water shrews and other small mammals employing plastic bucket traps. Their study area covered most of the Canadian range of the Pacific water shrew, although a limitation was that no sampling was done in regional or provincial parks.  At each location, their sampling design consisted of one trap line of 15 trap stations set along a watercourse (distance from water not given) and one or two other trap lines set 50-100 m from watercourses in forest. Traps were set at a location for two to five weeks. Their sampling effort of 13,462 trap nights produced only 3 Pacific water shrew captures at 3 separate locations.

Abundance

No estimates of population density exist for this species from any part of its range, and the total number of individuals and the number of mature animals in Canada is unknown. The Pacific water shrew appears to be rare throughout its range. Few captures of this species were taken in small mammal studies done in coastal Oregon and Washington (Aubry et al. 1991) and its relative abundance in various habitats was low compared with other shrew species. With its large body size and specialized feeding niche, the Pacific water shrew would be expected to be a rare species in shrew communities. A general pattern consistent among shrew communities in temperate regions (Churchfield 1991) is that the intermediate-sized species tend to be numerically dominant. The 8 captures of Pacific water shrew taken by Seip and Savard (1990) and Zuleta and Galindo-Leal (1994) despite an intensive sampling effort of 41,622 trap nights (Figure 5) can be attributed to both the difficulty of detecting this shrew even with effective traps such as pitfalls (see Bury and Corn 1987) and its natural rarity.  As an example of the difficulty in detecting this species and determining its absence, Zuleta and Galindo-Leal (1992) captured no Pacific water shrews at three separate locations in Clayburn Creek on Sumas Mountain in their 1992 survey. Yet a Pacific water shrew was accidentally captured in a Gee-type minnow trap set in Clayburn Creek in 1995 during a fish survey.

Fluctuations and trends

Nothing is known about yearly population fluctuations in the Pacific water shrew. Long-term population trends can only be inferred from anecdotal accounts and evidence based on the capture of museum specimens. The only published information on the status of mammals in the lower Fraser River valley in early historical time was the review by Allan Brooks (Brooks 1902). He described the Pacific water shrew as “fairly common” in the Chilliwack region and he collected a large series of museum specimens from this area in the late 1800s. Of the 114 Canadian museum specimens (collected from 1888-1999) that exist for this species, 29 (25%) were taken by Brooks in a 3-year period from 1895-1897 at “Sumas” a location that probably refers to the Sumas Prairie where according to Laing (1979) much of Brooks’ collecting was done in this period. The Sumas Prairie was a large wetland of marshes and sloughs associated with Sumas Lake that was drained by 1924 (Moore 1990). Given the primitive and inefficient traps available in the late 1800s, and the low capture rates of Pacific water shrew in the 1990s using more effective pitfalls traps (see previous section), Brooks’ capture of 29 animals in 3 years is noteworthy, suggesting the possibility of higher population densities in the late 1880s before large scale habitat change. 

Rescue effect

The Pacific water shrew inhabits Whatcom County in Washington adjacent to the Canadian border. According to John Fleckstein (pers. comm.) Washington has re-classified this species as S4 from S5 because of habitat change in the Puget Sound lowlands. The general habitat model in Johnson and Cassidy (1997) shows this species widespread adjacent to the Canadian border. However, because little is known about the actual distribution and population status of this shrew in areas adjacent to the Canadian population, it is difficult to evaluate the potential for rescue effect from Washington. It is noteworthy that the portion of the Canadian range adjacent to the Washington population is an area heavily impacted by habitat loss. Suitable habitat for immigrants would be minimal and highly fragmented.

Page details

Date modified: